Usefull Hearing Preservation Is Improved in Vestibular Schwannoma Patients Who Undergo Stereotactic Radiosurgery Before Further Hearing Deterioration Ensures

The present study evaluates whether hearing deterioration during observation reduces serviceable hearing preservation rates after stereotactic radiosurgery (SRS) in vestibular schwannoma (VS) patients with useful hearing. Methods We retrospectively analyzed 1447 VS patients who underwent SRS between 1992 to 2017. We identied 100 VS patients who had Grade I Gardner- Robertson (GR) hearing at initial diagnosis but were observed without surgery or SRS. We compared hearing after SRS in 67 patients who retained GR Grade I hearing from initial diagnosis to SRS (the hearing maintenance or HM group) to 33 patients whose hearing worsened from GR grade I to grade II (the hearing deterioration or HD group). We also investigated whether a decline in pure tone average (PTA) or speech discrimination score (SDS) before SRS affected hearing preservation after SRS.


Introduction
Vestibular schwannomas (VS) generally are slow-growing tumors that arise from the vestibular portion of the eighth cranial nerve. The most common presenting symptom is ipsilateral hearing loss [1]. The wide spread use of Magnetic Resonance Imaging (MRI) has led to earlier diagnosis of often smaller volume VS, so that some patients have remarkably good hearing at diagnosis [2,3]. For such patients observation or "wait and scan" is often recommended [4]. For patients with symptomatic larger volume VSs, microsurgical resection remains a potent intervention [4]. Stereotactic radiosurgery (SRS) for VS was rst reported in 1971 by Leksell [5]. SRS using the Leksell Gamm Knife (GK) (AB Elekta) for small to medium size VSs has a resection free rate of over 95% in 10 years with an associated transient facial neuropathy rate of 1-5% and a serviceable hearing preservation rate of 43-68% at 5 years after SRS [6][7][8][9]. Early SRS after diagnosis improves hearing preservation [10,11]. Based on referral patterns and wait and scan recommendations from initial consultants some patients with serviceable hearing were observed and underwent SRS only after hearing deterioration or tumor progression was recognized. The aim of this study was to evaluate whether hearing deterioration during observation of VS patients with GR grade I affected hearing preservation after SRS.

Patients population
We retrospectively analyzed 1447 VS patients who underwent GK SRS at our center between 1992 and 2017. Patients who underwent SRS before1992 received higher margin doses (18 Gy-20 Gy) and were excluded from this study. We also excluded patients who had undergone prior resection, prior radiation therapy, or who had neuro bromatosis type 2. We identi ed 100 VS patients who had Gardner-Robertson (GR) grade [12] I hearing at initial diagnosis and were then were observed. Serviceable hearing was de ned as GR grade I and II (PTA <50 dB and SDS >50%). The ow diagram of this study population is shown in Fig.1. We compared long term hearing after SRS in 67 patients whose hearing remained GR grade I from initial diagnosis to SRS (hearing maintenance or HM group) with 33 patients whose hearing worsened from GR grade I to II at the time of SRS (hearing deterioration or HD group). We also evaluated whether hearing preservation was related to worsening PTA and SDS values during the observation.

Radiosurgery Technique
SRS was performed in a single procedure that began with stereotactic head frame application using local anesthesia supplemented by intravenous conscious sedation. Patients then underwent high-de nition MRI and/or CT imaging studies as appropriate. SRS was performed using various models of the Leksell Gamma knife R (Model U, B, C, 4C, Perfexion and Icon (Elekta AB)). Dose planning was performed using various versions of the Leksell dose planning software (KULA or Leksell Gammaplan) [6,14]. Radiation

Statistical analysis
The data was analyzed using SPSS Statistics, version 25.0 (IBM, New York, USA). The relationship between HM and HD patients were analyzed statistically using Fischer's exact test and Mann-Whitney analysis as appropriate. Age, sex, tinnitus before SRS, SDS at diagnosis, target volume, margin dose, maximum dose, interval between diagnosis and SRS, and follow up time were not signi cantly different between these groups. We did nd that HM patients reported less vertigo symptoms and had better PTA values than HD patients (Table 1). More HM patients had Koos grade I tumors compared to HD patients (Table 1). The detailed difference between HM and HD patients are shown in Table 1. Tumor enlargement after SRS was de ned as a tumor volume that increased by ≥15%. Kaplan-Meier plots for tumor control rates were created. Hearing deterioration was de ned as a decline from GR grade I or II (serviceable) to III-V (non-serviceable, SDS <50% or PTA >50 dB). Kaplan-Meier plots for hearing preservation rates were created starting at the date of SRS. Univariate analysis was performed on the Kaplan-Meier method using a log-rank test. Multivariate analysis was performed using the Cox proportional hazards model. The suggested cutoff value for variables (age, PTA at diagnosis, SDS at diagnosis, target volume, margin dose, and interval between diagnosis and SRS) were determined by a Youden index based on receiver operating characteristic curve analysis [15]. P value of <0.05 were considered statistically signi cant.

Tumor control
The tumor control rate after SRS was 98.9% at 3 years, 95.3% at 5 years, and 92.2% at 10 years. In univariate analysis, there was no signi cant difference in tumor control rates between HM and HD patients (p =0.535). Four patients had tumor progression during follow-up. Three patients were observed without additional treatment because the tumors had initial enlargement but then stabilized. These patients were one HD patient and two HM patients, but all three patients had hearing deterioration from 12 to 54 months after SRS. A single patient underwent additional SRS (margin dose 11.0 Gy) at 4.5 years after initial SRS. This patient was HM patient, maintained useful hearing, and had no complications at 23 months after repeat SRS.

Serviceable hearing preservation
At the last follow up, 42 of 67 HM patients maintained serviceable hearing, whereas 8 of 33 HD patients had preserved serviceable hearing. The detailed comparison of GR hearing at diagnosis, SRS, and last follow-up are shown in Table 2. The overall serviceable hearing preservation rate was 65.9% at 3 years, 52.5% at 5 years and 40.5% at 10 years (Fig. 2a). The serviceable hearing preservation rates of HM patients were 79.9% at 3 years, 63.4% at 5 years, and 51.2% at 10 years. In contrast the serviceable hearing preservation rates of HD patients were 40.0% at 3 years, 32.7% at 5 years, and 19.6% at 10 years ( Fig. 2b). In univariate analysis, younger age (<55 years, p=0.006), better PTA at diagnosis (<20 dB, p=0.024), and HM status (p <0.001) were associated with improved long-term serviceable hearing preservation rates. The following factors were not signi cantly associated with serviceable hearing preservation rates: sex (p=0.119), tinnitus (p=0.643), vertigo symptoms (p=0.118), Koos  During the observation interval (median interval 17.4 months) before SRS, the median PTA increased from 16.9 dB to 21.25 dB and the median SDS decreased from 96% to 92%. The median increase in PTA was 6.85 dB and the median decrease in SDS was 4%. In univariate analysis, patients with an increase in the PTA ≥15 dB during observation had reduced serviceable hearing preservation rates (p=0.024) (Fig.   2c). Similarly, patients whose SDS decreased ≥10% had reduced serviceable hearing preservation rates (p=0.024) (Fig. 2d). GR grade I HM patients whose PTA worsened by ≥6 dB had signi cantly reduced GR grade I maintenance (p=0.005) (Fig. 2e). Patients whose SDS decreased by ≥5% also had a reduced probability of maintaining GR grade I hearing (p=0.028) (Fig. 2f). Patients with a PTA at SRS of <20 dB (p=0.010) had a higher probability of maintaining GR grade I.

Discussion
Current management options for newly diagnosed VSs include observation, SRS, and surgical resection [4,17]. In prior decades before the wide availability of MRI for screening of patients with asymmetric hearing loss, tinnitus, or disequilibrium symptoms, many VS tumors were not detected until the tumor resulted in major cranial nerve or neurological dysfunction. For symptomatic larger tumors with brain stem compression early complete resection became the most frequent intervention in patients without signi cant medical comorbidities. In patients with signi cant medical comorbidities subtotal resection followed by adjuvant SRS more recently has become an additional option [18][19][20]. In 1971, Leksell rst reported the potential role of SRS for vestibular schwannomas (acoustic neuromas) [5]. In 1989, Kondziolka et al reported 162 consecutive VS patients who underwent Gamma knife SRS; they noted tumor control in 98% and unchanged hearing in 51% of patients [21]. In 2006, Pollock et al. reported a prospective cohort study comparing surgical resection with SRS for patients with small-to moderatesized VS [22]. They described no difference in tumor control, and described excellent cranial nerve preservation rates. After these and many other SRS reports emerged, further observation of smaller tumors rather than early surgical intervention was increasingly recommended.
Further observation was based on the premise that VS tend to grow slowly, and in any case outcomes from intervention are not worsened during an observational interval after the rst con rmatory MRI [23,24]. The present report examines the question of whether even mild hearing deterioration during an initial observation (or "wait and scan") interval leads to worse hearing outcomes in a series of patients who eventually proceed with SRS.

Prior Reported Hearing Results in SRS Patients
We reported that VS patients with serviceable hearing at the time of SRS had overall serviceable hearing preservation rates of 77.8% at 3 years, 68.8% at 5 years, and 51.8% at 10 years [14]. Akpinar et al. compared hearing outcomes in early SRS patients (<2 years after diagnosis) with late SRS hearing outcomes (>2 years after SRS). This report con rmed that earlier SRS resulted in better hearing preservation outcomes [11].
In the present study we evaluated hearing outcomes in patients who had normal or GR Grade 1 hearing at the time of initial diagnosis. In this retrospective study some patients maintained GR 1 hearing during the observation interval until SRS was performed. We compared those patients to a similar group of patients whose hearing had worsened during the observation interval. The multivariate analysis revealed that those without hearing worsening (HM Patients) and those younger than 55 years had signi cantly better useful hearing preservation rates after SRS. We and others have previously reported the bene cial effect of younger age in hearing preservation results after SRS [9,14]. The present study con rms that only those patients whose hearing does not worsen during observation have the best chance of long-term hearing preservation once SRS is performed. Patients whose hearing deteriorated during observation before SRS had worse hearing outcomes after SRS.
Regis et al. performed a study that evaluated intracanalicular VS patients who underwent early SRS or "wait and scan strategy" [10]. They found that tumor control and functional hearing preservation rates were higher in patients who underwent early SRS.
Many patients are now diagnosed by MRI performed for relatively mild or even non-speci c symptoms such as asymmetric hearing loss, vague imbalance symptoms, vertigo, or tinnitus. If a small VS is found on MR, these patients are often referred to an otolaryngologist or neurosurgeon. After review of the hearing results, they may recommend a period of observation with a new scan in 6 -12 months. Patients may be told their tumor is benign and grows slowly. Some patients delay subsequent imaging if the symptoms remain mild and non-progressive. Referral for intervention by either surgery or radiosurgery may not occur for months or even years.

What We Learned
Based on our experience in more than 2000 VS patients over a 32 year interval, we suspected that earlier SRS might lead to better hearing preservation rates-and that hearing maintenance is an important outcome measure for many VS patients. We found that both PTA and SDS measurements are important predictors of hearing outcomes during such an observation interval. VS patients whose PTA increased >6 dB or whose SDS decreased by >5% were less likely to retain GR grade I after SRS. Even more dramatic changes such as an PTA increase of >15 dB or a SDS decrease >10% during observation led to less serviceable hearing preservation. Of interest, Kirchman et al reported 10 years hearing outcomes in observed intracanalicular VS patients [25]. They found that the median PTA increased from 51 dB to 72 dB and median SDS decreased from 60% to 34% after follow-up of 9.5 years. In the current experience series, the median PTA increased from 16.9 dB to 21.3 dB at the time of SRS (median interval 17.4 months) and SDS decreased from 96% to 92%.

Study limitations
This is a retrospective study, and lacks data related to tumor volume changes during the interval between rst diagnosis and SRS. The HM group and HD group were not matched for PTA at diagnosis, presence of vertigo before SRS, and Koos grade. This study found that even mild hearing deterioration during the observation period before SRS signi cantly reduced the ability to maintain useful hearing after SRS.

Conclusions
To improve the long term maintenance of serviceable hearing in VS patients with Grade 1 hearing at diagnosis, we advocate SRS before further hearing deterioration is detected.